Files
ray/python/ray/rllib/tuned_examples/humanoid-ppo.yaml
T
Eric LiangandGitHub 995ac24a2c [rllib] clarify train batch size for PPO (#2793)
It's possible to configure PPO in a way that ends up discarding most of the samples (they are treated as "stragglers"). Add a warning when this happens, and raise an exception if the waste is particularly egregious.
2018-09-05 12:06:13 -07:00

18 lines
370 B
YAML

humanoid-ppo:
env: Humanoid-v1
run: PPO
stop:
episode_reward_mean: 6000
config:
gamma: 0.995
kl_coeff: 1.0
num_sgd_iter: 20
lr: .0001
sgd_minibatch_size: 32768
train_batch_size: 320000
model:
free_log_std: true
use_gae: false
num_workers: 64
num_gpus: 4